TUNICATA OF THE PEIBILOF ISLANDS. 531 



At least 17 series of stigmata, the individual stigmata exceedingly small and 

 obscure, though the series fairly well marked by the heavy muscle band in each inter- 

 stigmatic space. This muscle band as broad as, or broader than, the stigmatic area 

 (flg. 28, m. c. V). 



Bndostyle broad, not greatly tortuous. No information concerning the dorsal 

 languets. 



Digestive apparatus. — Loop rather wide, not twisted, particularly characterized 

 by the great thickness of the rectal limb. 



OEsophagus exceptionally broad, particularly at its mouth ; much narrower at its 

 entrance into the stomach. Stomach apparently nearly spherical in its normal form, 

 but usually broader than long in preserved specimens. Its entire wall covered with 

 small, rather uniform, thickened patches or areolations (flg. 28). Length of the loop 

 behind the stomach considerably greater than cambined length of oesophagus and 

 stomach. Eectal portion of intestine greatly enlarged in most specimens. It reaches 

 the dorsal side of the branchial sac by turning at a short angle immediately behind 

 the sac. 



Reproductive organs, — Situated in the long," broad post-abdomen. Ovary in the 

 form of a long, narrow band placed at one side of the post-abdomeu (fig. 28, ov.), the 

 ova distinctly amoeboid, and without recognizable follicular epithelium or "test" cells. 

 No oviduct observed. Testis distinguished with difficulty (excepting 

 when containing ripe sperm) from the great mass of mesenchymatous jj j 



material by which the post-abdomen is filled. Vas deferens not seen. 

 Embryos develop in packages in cavities of the test (fig. 27). No bud- 

 ding seen. 



(See end of paper for account of sexual reproduction of this species.) 



This species is certainly closely related to S. turgens Phipps, and 

 at first I was much inclined to consider it to be identical with that spe- 

 cies. There seem, however, to be several points of difference between 

 them so considerable as to make it necessary to regard them as distinct species. 



In the first place 8. turgens as figured by both Savigny, 1816 (PI. Ill, fig. 3), and 

 Sars (see Bonnevie, 1896, p. 12, and PI. lY, fig. 36), has the lobes in general much 

 more separate than is the case in 8. irregulare. And in his description Savigny 

 speaks of the systems as being in the form of solid cylinders "isolated or associated by 

 their peduucles." Again, our species shows no trace of the channels on the surface of 

 the lobes marking the intervals between the zooids, such as are described and figured 

 by Savigny in 8. turgens. The systems and common cloacal orifices of 8. turgens are, 

 according to the authors already mentioned, very distinct, while in 8. irregulare, as will 

 be noted from my description, there is much doubt about their being present at all. 



As concerns the zooids, it would appear that the two forms agree very closely, 

 though it is hardly possible that the branchial sac of 8. turgens could be so closely 

 contracted, and the stigmata thereby so greatly obscured, as is the case in all the 

 material of 8. irregulare at my disposal, without having been mentioned by Savigny. 



From 8. incrustatum, Sars, Huitfeldt-Kaas, 1896, the only other species of the 

 genus, the present species differs so markedly that a detailed contrast between them 

 would be superfluous. 



The collection contains about a dbzen colonies and pieces of colonies of this most 

 interesting ascidian, all, so far as my information goes, from St. Paul Island. 



